Re: Re[2]: bbdb-initialize in .emacs
[EMAIL PROTECTED] (Michael A. Miller) writes: > I'm trying to set up bbdb and am unclear on the process for the > bbdb package. The docs (/usr/share/doc/bbdb/README.Debian) say > that to enable bbdb support, I need to add a call to > bbdb-initialize in my .emacs. So I added I presume someone already got back to you, but you probably also need a (require 'bbdb) before the initialize statement in your .emacs. Bbdb doesn't appear to do this automagically (it's commented out in the /etc/emacs20/site-start.d/50bbdb-init.el file), and that's good thing because it makes it possible for people to choose whether or not they want bbdb, and it also makes it much easier for users to install/use local versions. Hope this helps. -- Rob Browning rlb @defaultvalue.org, @linuxdevel.com, and @debian.org Previously @cs.utexas.edu GPG=1C58 8B2C FB5E 3F64 EA5C 64AE 78FE E5FE F0CB A0AD
Re: emacs 20.5a uploading (Possibly important Y2K fixes).
[EMAIL PROTECTED] (Thomas Bushnell, BSG) writes: > Did you miss Miles Bader's correct message about emacs version > numbering? No. His email prompted my question. I didn't realize he was *sure*, and I wanted to double check before I go make irreversable (well without using epochs) changes in the Debian package. The real problem was that I didn't notice the release of 20.5 until 20.5a was available. If I had seen 20.5, then I wouldn't have been confused. I must not be on the right emacs list for announcements... > Emacs 20.5a IS newer than emacs 20.5. Thanks. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
emacs 20.5a uploading (Possibly important Y2K fixes).
I'm very happy that there weren't any serious upstream conflicts to integrate so I was able to package it quickly. It's uploading now, but I'm on a slow connection ATM, so it might be a while... Although I've been told there are some y2k fixes in this version, I haven't carefully evaluated the diffs. This is just an upstream bug-fix release, so no new features. If someone's willing and able, this should probably be considered for the slink-y2k package set too. Let me know if there are any problems. I'm travelling at the moment, so my access may be a bit irregular. Thanks P.S. The version number is 20.4.pre20.5a-1. This avoids the problem with the fact that the upstream tarfile's version 20.5a sorts (via dpkg) as newer than 20.5 which hasn't been released yet. Epochs would be another solution, but I haven't decided whether or not I like that better, and the packaging manual seems to lean against it. FWIW -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: LogCheck and it's rules
"Paul J. Keenan" <[EMAIL PROTECTED]> writes: > The logcheck script is in /usr/sbin/logcheck.sh - the script uses > grep to do the pattern matching. From the source and the grep(1) > manpage, it seems that for the lines to include in the log > (logcheck.hacking and logcheck.violations) the matching is > case-insensitive, but for the exclusions (logcheck.violations.ignore > and logcheck.ignose) the matching is case-sensitive. HTH. Actually, the Debian package uses egrep. Check the script. That's why (as you correctly stated) you need \[. I filed bugs against the docs and against the included default patterns a while ago, and I believe it's being fixed. For example, here are some (correct?) patterns I added: uservd\[[[:digit:]]+\]: call connected$ uservd/check\[[[:digit:]]+\]: uservd\[[[:digit:]]+\] is running$ named\[.*\]: Cleaned cache of .* RRsets named\[.*\]: USAGE .* .* CPU=.*/.* CHILDCPU=.*/.* named\[.*\]: NSTATS .* .* A=.* PTR=.* =.* named\[.*\]: XSTATS .* .* RR=.* RNXD=.* RFwdR=.* RDupR=.* RFail=.* RFErr=.* RErr =.* RAXFR=.* RLame=.* ROpts=.* SSysQ=.* SAns=.* SFwdQ=.* SDupQ=.* SErr=.* RQ=.* RIQ=.* RFwdQ=.* RDupQ=.* RTCP=.* SFwdR=.* SFail=.* SFErr=.* SNaAns=.* SNXD=.* -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
ATTN: stalin users (slightly OT)
Apologies for the slightly off-topic message, but I know of no other way to reach the Debian stalin users. The upstream author Jeffrey Siskind has asked if I could ask those who find stalin useful to send him a message stating such. He needs feedback to use during his evaluation to convince his employers to allow him to keep spending time on this project. Taking a few moments to contact the him will translate fairly directly into corporate support for a free software project if enough relevant people do it. Thanks Here's specifically what he asked: From: Jeffrey Mark Siskind <[EMAIL PROTECTED]> Subject: Re: stalin: patch for sparc compile. To: [EMAIL PROTECTED] Date: Mon, 27 Sep 1999 21:42:33 -0400 Reply-to: [EMAIL PROTECTED] Thanks for the patch. It looks OK to me. BTW, is there any way to get a measure of how many people are using Stalin through Debian? I'm going through my evaluation now and it would help me if I could say to my committee something like `at least n people are using Stalin as obtained through Debian'. Even better would be feedback from users. One or two line statements are fine. Something like `I use Stalin to support my research in widgets. I find that for my task, code compiled by Stalin runs x% faster than code produce by Foo-C, the previous Scheme compiler that I was using. Without Stalin, I would not be able to conduct my research.' Jeff (http://www.neci.nj.nec.com/homepages/qobi) -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Minor inittab/serial console question...
[EMAIL PROTECTED] (Miquel van Smoorenburg) writes: > I know what you want. You want to run "minicom" in a "screen" session. > Start it once, detach it whenever you want, attach it later. Hmm. Interesting. I've never used screen. I'll have to see how it works. I was worried about security, but if screen forces you to authenticate whenever you attach to a session, then that might work well. Would there be any reason I wouldn't want to have this session started at boot time as some desginated user so that it's always capturing that serial port's output? Also, I wonder if minicom can be configured to log all the I/O during a run... I'll go see when I get a chance. Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Minor inittab/serial console question...
[EMAIL PROTECTED] (Miquel van Smoorenburg) writes: > In article <[EMAIL PROTECTED]>, > Paul McHale <[EMAIL PROTECTED]> wrote: > >Can you boot Linux without a video card installed ? I would have thought > >the BIOS would have had a problem with that ... > > Yup, works just fine. In fact, if you have the serial console support > compiled in, the kernel will detect that you don't have a VGA console > and automatically use /dev/ttyS0 as console. And for the final touch, you can tell lilo about the serial port and have it use that too. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: emacs or xemacs ?
"Richard E. Hawkins" <[EMAIL PROTECTED]> writes: > also, using emacs on CCIE (*Control Key In Exile) keyboard, causing > enough long reaches on my little finger sthat I had to see a quack, > is probably part (put a minor part) of my avoiding either. You probably know this, but you can easily remap caps-lock etc. to be control... Also, you can get a better keyboard for about $3 at goodwill here. Of course that still won't help you if you prefer vim ... nothing will :> (holy-war-aversion-disclaimer: Yes, that was a joke, and yes, I prefer emacs, but no, I don't think it's for everyone, and yes, I have used vi (minimally), and can understand the appeal.) FWIW -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: sb live card under debian
"Jim Ruby" <[EMAIL PROTECTED]> writes: > Is there a driver for the sound blaster live card? > I have the sb.o module installed but don't quite get how to get it started, > it says i/o icq and dma is required. > > so I'm not show to do the i/o as it don't like i as an argument. > > modprobe i=6500 o=651f irq=9 dma=1 As far as I know the SBLive is completely unusable under Linux, and Creative Labs is refusing to give any programming info for the card so there may not be a free-software driver in the near future. But you can get good PCI cards that are supported for about $20 now, so it's not that big a deal. See www.alsa-project.org for one list of supported cards. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Minor inittab/serial console question...
Seth R Arnold <[EMAIL PROTECTED]> writes: > Rob, why not do everything via telnet or ssh? two network cards can be had > for under $50 if the machines don't already have network cards.. The serial console is to handle cases where the machine can't even boot to the telnet or ssh stage. This machine is critical, so even if it goes down when I'm away, I want to be able to dial up another machine on it's local net and reach it's console via the serial port. Then I can fix whatever's wrong (even lilo problems) and reboot it (via X10 power cycling). It's also just nice to be able to watch the machine's boot messages now and then without having to put a video card in there to make sure everything's OK. > (sorry for the lack of real response, I don't know how to do what you > want... so I will try to find out why you want it. :) Well, now that I think about it, I'm not sure I want what I thought I wanted. I wasn't thinking too clearly last night (too much time spent setting up a mess of stuff). What I'd really like is some little daemon on the machine connected to the headless machine that watches and logs the serial port output, and that you could connect to whenever you wanted (access would be exclusive) to communicate with the other machines console and to page back through the log. That daemon would "own" the port. Sort of a virtual virtual terminal... For now I get 90% of what I want just using minicom, so I'll probably just stick with that... Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: script command question
Mark Wagnon <[EMAIL PROTECTED]> writes: > Cool. Thanks. I needed an example of how to get the script to > receive the text to be filtered. This'll help a lot. Thanks again. You might also be interested in the filterm program from the konwert package. I came across that today and from a brief inspection, I recall that it may do what you want. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Minor inittab/serial console question...
I've set up one of my headless machines to use ttyS0 as a serial console, and I've also set it up to run getty on that port after it finishes booting. Using minicom from another machine, I can monitor the other machine's serial port and see that everything works fine. After it boots, I can log in, etc, and I can even interact with LILO using it's "serial=" option, but now I'd like to set it up so that the non-headess machine automatically puts its ttyS0 (the one connected to the headless machine's serial console) on one of its virtual terminals. Is there an easy way to do that? I've hunted around the HOWTOs and linux/Documentation/* and I can't find anything relevant. Any help or RTFMs would be appreciated. Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: How to remove a user
David Teague <[EMAIL PROTECTED]> writes: > I think you issue the command > > userdel -r username > > to remove both user, the home directory and files, but RTFM > to be sure. Note that this won't get any files the user might have in locations other than their home directory, and I don't know if it deletes their group (though it probably does). To find and delete all the files the user owns, you'll need something like this (as root): find / -name username -o -group usergroup > /tmp/usersfilelist Then you can delete all the files in usersfilelist. This probably won't catch things like the user's crontab or any "at" tasks. You'll have to handle those manually, and to be truly safe, you need to lock the user out before doing this (I've forgotten the magic command for that, but you can just change their password with passwd and use who or w to make sure they're not currently logged in). Hope this helps. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: ldap/pam problem
Jason Gunthorpe <[EMAIL PROTECTED]> writes: > I used this complex invokation, you'll need an appropriately bug-fixed pam > library (Ben, you have my patches..) That's nice. Thanks. That solves at least part of my problem, though I'm beginning to think that we may be better off cobbling up something here that's more tailored to our needs and quit spending so much time dealing with fairly bleeding edge stuff. I may give it one more go tomorrow... Thanks again. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
ldap/pam problem
Before I ask more detailed questions, I wanted to know if I really need to edit the /etc/pam.d files presuming that I've modified my /etc/nsswitch.conf file? If so, then when I tried those modifications, I couldn't figure out how to get reasonable behavior. If you have auth sufficient pam_ldap.so auth required pam_unix_auth.so try_first_pass then if the entry is found in ldap, pam returns and you never execute things like motd, etc. which is not what you want. Presuming that setting up /etc/nsswitch.conf is sufficient, I still can't get things to work. If I try to log in as a normal (non-ldap) user, everything's fine, and I can see that it's asking the ldap server for the user and getting no answer (as an aside, it looks like it's asking the ldap server for a bunch of fields that the MigrationTools didn't create (shadowmax, shadowmin, etc)). Then it lets that user log in as usual based on their shadow passwd entry. However, if I try to log in as stray (which is the user that has no shadow passwd/group entry, but has an ldap entry), it never even asks the ldap server, it just fails. If I try su instead, I get "su: problem establishing PAM_RUSER" also with no attempt to contact the ldap server. I think I got my pam_ldap.conf right -- the relevant lines should be: base ou=People,dc=localnet pam_filter objectclass=account My nsswitch.conf reads as indicated in the libnss-ldap readme: passwd: files nis ldap group: files nis ldap shadow: files nis ldap (I didn't know about shadow, but I presumed it should be the same as the others). And I can tell that the entry for stray is in the ldap database because I can see it with ldapsearch: $ ldapsearch -h localhost -v -b ou=People,dc=localnet "(& (uid=stray) (objectclass=account))" ldap_init( localhost, 0 ) filter pattern: (& (uid=stray) (objectclass=account)) returning: ALL filter is: ((& (uid=stray) (objectclass=account))) uid=stray,ou=People,dc=localnet uid=stray cn=stray objectclass=top objectclass=account objectclass=posixAccount loginshell=/bin/bash uidnumber=1008 gidnumber=1008 homedirectory=/home/stray gecos=,,, 1 matches Any help would be appreciated. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Some email bounces = misconfigured exim?
George Bonser <[EMAIL PROTECTED]> writes:
> This is also a correct method. By setting your HELO to the name your IP
> address resolves to, you are assured of delivery to most every site that
> accepts mail from ISP dialups.
Right. I get the same effect (though there's a minor bug because DNS
isn't always happy as soon as the IP address changes) from my dhcpcd
script with exim like this:
#!/bin/sh
source /etc/dhcpc/dhcpcd-eth1.info
logger echo "`date` dhcpcd-sv: IP address changed to $IPADDR"
RR_NAME=`host ${IPADDR} ${DNS} | grep Name | cut -b 7-`
# Tell exim what happened.
(cd /etc && \
perl -pi -e "s/^RR_HOSTNAME = (.*)/RR_HOSTNAME = ${RR_NAME}/o" exim.conf)
exit 0
And I just have this in my exim.conf file:
# This is who we want to be known as. It also affects the outgoing
# HELO messages, so you want it to match exim's outgoing IP
RR_HOSTNAME = cs2868-35.austin.rr.com
primary_hostname = RR_HOSTNAME
etc.
> Note that many mailhosts not only refuse mail from blocks of IP
> addresses that are known dialups but also reject mail from
> hostsnames including such patterns as "cust" or "dialup" or
> "dynamic"
Yep. I haven't really had a problem (that I know of) with this, but
it's a choice between two evils. If I do this, I risk being shut out
by people that block dynamic ip's, but if I use the RR smtp relay,
then I'm relying on RR to keep their server up and well maintained and
that doesn't give me a warm fuzzy feeling either...
--
Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Further work on LDAP passwords (working on an ldap-adduser).
[EMAIL PROTECTED] writes: > Yup, it's at http://www.umich.edu/~dirsvcs/ldap/doc/guides/slapd/ Thanks. I don't know how I overlooked that. I noticed in the README for libpam-ldap that you need to use some secure socket mechanism if you really want an ldap setup to be secure. Is there a doc somewhere that explains the best way to set that up? The README mentions several alternatives (CRAM-MD5 and SSL/TLS), but I don't know enough to know which would be preferable or how to set them up. Also, I'm wondering if it would be useful (if it hasn't been done already) to generate chsh and chfn replacements (like the passwd one) to handle changing the attributes in the ldap server rather than locally when the user's info is in ldap. I suppose this would require augmentation of the pam-apps package... Is there any overriding plan to integrate all this stuff? What would be nearly ideal is if there were a config file somewhere where you could just tell the system that all user accounts should be handled by ldap and have all the appropriate tools do the right thing. This looks like the direction things are going with libpam-ldap, but I didn't know if that was the final goal. Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Further work on LDAP passwords (working on an ldap-adduser).
[EMAIL PROTECTED] writes: > Yea, this one stumped me for quite a while to. From section 5.3 of > the Slapd administrators guide: NOTE: The DN pattern specified > should be "normalized", meaning that there should be no extra > spaces, and commas should be used to separate components. An example > normalized DN is "cn=Babs Jensen,o=University of Michigan,c=US". An > example of a non-normalized DN is "cn =Babs Jensen; o=University of > Michigan, c=US". Is the Slapd administrator's guide available on-line, or is that the actually published book you're talking about? > So you need to get rid of the spaces in your access line. i.e.: > > access to * by dn="cn=admin,ou=People,dc=localnet" write Ooh, nasty. This was exactly the problem. Thanks so much for the help. No one would have wanted to be around me by the time I would probably have figured this out myself... > I'll go ahead and report this to the openldap maintainer. Great, thanks. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Further work on LDAP passwords (working on an ldap-adduser).
Jason Gunthorpe <[EMAIL PROTECTED]> writes: > You use the same technique as ethernet basically, both add and check > if their was only one added (using a search on the ID they added) > then remove and retry a new id after a delay. Oh, right, of course. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Further work on LDAP passwords (working on an ldap-adduser).
Jason Gunthorpe <[EMAIL PROTECTED]> writes: > You don't really need this, just a simple detect/backoff algorithm will do > OK for determining the UIDs Could you elaborate? If you have at least two machines, I can't see how you don't have to have a shared locking mechanism to make sure they don't both try to add the same user to the ldap database with different uid's, or are there some ldap semantics that allow you to avoid the race condition. > > 2) get the next user id (how?). [1] > > The only way I could see is to suck down the entire UID list and pick the > highest, it is a big search, but ldap doesn't have a mechanism for > integer value compares so it is the best I could think of. It won't be that big here. Ldap's only going to be managing the local users, and there will only be a handful or two. And if I just mandate that all account maintenance (add/delete/modify) be handled from a single machine, I could just store the last uid/gid in a local state file. This would also avoid the problems in the previosu point above. > > 5) Run a script to do whatever's needed to create the user's home > > directory in all the right places [1]. > > The pam_mkhomedir module I cooked up is what I think we will ultimately > use on debian.org, the home dirs just spring into existance on the first > login. Saves resources+headaches Right, but that doesn't help if you need to do something more complicated (or even just different). It might be nice to modify the mkhomedir module to use a config file, and just have that config file allow you to specify the script to be run to handle things. The default could be something like /sbin/pam-mkhomedir which would have the current behavior. Of course, maybe it just makes sense for us to take pam-mkhomedir and modify it for our purposes. Here we really just need it to make sure the user's CODA volume is available and ready. I'm wondering if there's anything I'm going to have to do here that could be generalized and useful to others, or if it's just going to be too domain specific to be useful to anyone else. (Oh, and I guess you didn't have any idea why my script was failing with authentication problems?) Thanks for the help. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Further work on LDAP passwords (working on an ldap-adduser).
OCALNET" ) slapd[24001]: > cache_find_entry_dn2id: found dn: OU=PEOPLE,DC=LOCALNET slapd[24001]: <= dn2id 2 (in cache) slapd[24001]: => id2entry_w( 2 ) slapd[24001]: > cache_find_entry_dn2id: found id: 2 rw: 1 slapd[24001]: entry_rdwr_wtrylock: ID: 2 slapd[24001]: <= id2entry_w 0x808ca30 (cache) slapd[24001]: => access_allowed: entry (ou=People, dc=localnet) attr (children) slapd[24001]: => acl_get: entry (ou=People, dc=localnet) attr (children) slapd[24001]: => acl_get: edn OU=PEOPLE,DC=LOCALNET slapd[24001]: => acl_get: [1] check attr children slapd[24001]: => dnpat: [2] .* nsub: 0 slapd[24001]: => acl_get:[2] backend ACL match slapd[24001]: => acl_get: [2] check attr children slapd[24001]: <= acl_get: [2] backend acl ou=People, dc=localnet attr: children slapd[24001]: => acl_access_allowed: write access to entry "ou=People, dc=localnet" slapd[24001]: => acl_access_allowed: write access to value "any" by "CN=ADMIN,OU=PEOPLE,DC=LOCALNET" slapd[24001]: <= check a_dnpat: CN=ADMIN, OU=PEOPLE, DC=LOCALNET slapd[24001]: => string_expand: pattern: CN=ADMIN, OU=PEOPLE, DC=LOCALNET slapd[24001]: => string_expand: expanded: CN=ADMIN, OU=PEOPLE, DC=LOCALNET slapd[24001]: => regex_matches: string: CN=ADMIN,OU=PEOPLE,DC=LOCALNET slapd[24001]: => regex_matches: rc: 1 no matches slapd[24001]: <= acl_access_allowed: denied by default (no matching by) slapd[24001]: => access_allowed: exit (ou=People, dc=localnet) attr (children) slapd[24001]: no access to parent slapd[24001]: send_ldap_result 50:: slapd[24001]: > cache_return_entry_w slapd[24001]: entry_rdwr_wunlock: ID: 2 slapd[24002]: do_unbind Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: We need centralized accounts -- Any docs for ldap passwords?
Sergey V Kovalyov <[EMAIL PROTECTED]> writes: > libpam-ldap will allow password change. The rest have to be done > manually (or through some customized software. I am considering > Ganymede.) Although there is a nice package pam-mkhomedir that will > automatically create homedirs (and copy /etc/skel stuff) if it does > not exist. Where can I find that script? I think we're going to go with ldap, so I'm going to have to figure out how we want to handle adding/deleting users, etc. I suppose I'll just whip up some scripts, but I wouldn't mind having a good one as a reference. First I need to look in to ldap and see what the tools are for editing the database from the command line (if that's possible). Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: We need centralized accounts -- Any docs for ldap passwords?
Ben Collins <[EMAIL PROTECTED]> writes: > Documentation is a little lacking in this area. The main reason for > putting things like fstab, etc, into ldap is for diskless clients > and large network configurations (think centralizing). If you don't > see an immediate need for it, chances are you wont benefit from > it. Currently the most common use of ldap for name services are > shadow/passwd/group, mail aliases (exim can compile with ldap > support, as well as sendmail), and hosts information. OK, so it sounds like we just need shadow/passwd/group support, and as far as I can tell we should be mostly good to go if we 1) firewall access to the ldap server from outside our subnet. 2) import etc/group and passwd via migrate_.pl 3) edit our nssswitch.conf as directed in /usr/doc/libnss-ldap/README 4) cross our fingers. What I don't really know is how doing this interacts with the normal mechanisms. I would presume that we can just use LDAP for user accounts, and leave the system accounts in /etc/passwd, etc. I'm guessing from the nsswitch entry it'll just fall back to that if LDAP fails on a given lookup, but how does LDAP interact with adduser, userdel, addgroup, /usr/bin/passwd, etc. Does it update the right things, or do we have to do manual synchs? If the latter, then it seems like it might be worth us considering not using LDAP at all, and just whipping up some ssh synch thingy for these bits... > Hope this clears some things up. It helps a lot. Thanks. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: We need centralized accounts -- Any docs for ldap passwords?
Sergey V Kovalyov <[EMAIL PROTECTED]> writes: > When you install libnss-ldap, there is a short howto in > /usr/doc/libnss-ldap > I also suggest downloading conversion tools from www.padl.com, which will > help populate the LDAP database OK. I'm back working on this, and I've gotten openldap etc. installed, and I've gotten the migration tools, read the HOWTO, and played with gq to see that I can actually see my database, and I'm about ready to try and cram my passwd/group stuff in there. However, from looking at the migration tools, it seems that they can translate a lot more than just passwd/group stuff like services, protocols, aliases, fstab, etc. So I'm a little curious now. I'd like to get a brief overview of the overall picture. Are people using ldap much for things like fstab? If so, how would that actually work, and how would it interact with other package upgrades? (I can see how accounts work via glibc2 and libpam-ldap/libnss-ldap.) Also, I'm wondering what, if any, the security concerns are relating to ldap access to passwd etc. Can someone give me a brief overview or point me at an appropriate doc? I haven't found one yet. Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: mail clients
Stephen Pitts <[EMAIL PROTECTED]> writes: > See /usr/doc/exim/filter.txt.gz > It answered all of my questions. See also "info exim-filter". -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Emacs and suidunregister
[EMAIL PROTECTED] writes: > I read the man page for suidregister, and it seems to me that when I > purged Emacs19 and Emacs20 (and all related binaries), not > everything was cleaned-up. It also seems to me that this might be a > bug. > > Can anyone confirm, or suggest what I did wrong? (I don't think it > will be any problem to correct.) This was a bug in the way I was calling suidunregister in the emacs20 postinst. I *think* it's fixed in the newer packages. If you have a reasonably inexpensive net connection, try installing the current emacs20 and then purging it. If that doesn't fix the problem, let me know. If you pay for downloads by the byte, then let me know and I'll investigate it here first. Thanks. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: ssh @ pseudo ttys
Ian Keith Setford <[EMAIL PROTECTED]> writes: > I just installed a new box but when I try to ssh in I recieve this > error: > > Warning: Remote host failed or refused to allocate a pseudo tty > > Thisonly happens with sshas I can telnet in and out with no > problems. Anybody know where the problems is? Or how I can turn > off the pseudo ttys altogether? My experience was that I had to recompile my kernels with psuedo-tty support (this meant that I had to have a new enough kernel). This was a fairly aggravating problem for one remote install. It might be nice if the preinst warned about this, especially if it can check in /proc to see if ssh is likely to be hosed after the install. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: What to do with a tape drive?
Carl Mummert <[EMAIL PROTECTED]> writes: > The easiest way to do incremental backups is to use a prewritten > package. I used 'tob' for a while, and it worked fine, although you > may have to spend an hour or two configuring it the first time. Also well worth investigating is amanda (www.amanda.org). It's probably more complex than most single machine users need, but it's certainly a solution you're unlikely to outgrow. I use it both at home and at work. However, it does require that you be willing to accomodate its model... -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: mailers
Brian Schramm <[EMAIL PROTECTED]> writes: > alias support for addresses > grouping of addresses > threading of email (for lists) > some kind of sorting on reciept of a massage ( for news reports) > easy to use > reliable file transfer (for attached documents) > find a message on a search of the header or body > > Let's here your favorite that can handle most of the above if not all. I'd recommend Gnus if you like emacs, and mutt if you don't. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: We need centralized accounts -- Any docs for ldap passwords?
"Jens B. Jorgensen" <[EMAIL PROTECTED]> writes: > But if you're not ready for potato then NIS will provide a > ready-made solution. It's pretty straightforward. I'd be glad to > offer assistance. As for a comparison, well, they're different. NIS > has been around a long time, LDAP is newer. Thanks to both of you for the help. We're running all unstable here so getting the ldap packages isn't a problem, but I wasn't completely sure what the tradeoffs are. I'll go look at the web sites Ben mentioned. Actually I had already seen the ldap pacakges, but I wasn't quite sure where to start. I was hoping for a HOWTO or something, but I can always just jump in and figure it out as I poke around. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
We need centralized accounts -- Any docs for ldap passwords?
We've got a number of machines here that we need to switch to centralized account maintenance, and I was trying to figure out what the best solution would be. It looks like the two main solutions would be NIS or ldap (via PAM), but I'm having a hard time finding out enough about the ldap solution to do a good comparison. Is there a good HOWTO or similar somewhere? Is there some other solution I've overlooked. (I thought about just using a cron job and a sync script to keep all the passwd/group files in sync, but that requires you to be able to atomically update the files, and I couldn't see a good way to do that...perhaps some trick with chpasswd/add/deluser... Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Partition hosed. Is there a way to find the superblock?
I did something stupid (not rebooting after a re-partition), and so now after I finally did reboot, I've got a partition with a bunch of data on it that can't be mounted. Both mount and e2fsck complain about not being able to find the superblock. I can get some of the data off of it by just running strings on the partition, but I'd really rather find the superblock and bring it back to life. The data there is not critical (it's just my Debian mirror partition), but I'd rather not have to reconstruct that through my modem link if I don't have to. Is there some "magic number" or unique header I can search for that will tell me where one of the (backup) super blocks is stored? Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930
Re: Bug reporting proceedure, was Re: Bug#24066: libc6: rsh segfaults as , a result of new libc 2.0.7r2
Manoj Srivastava <[EMAIL PROTECTED]> writes: > Make bug reporting any more onerous than it is, and peole > merely stop filing reports. I suppose there is something to that. > For the most part the maintainer knows the bugs on a package > better than anyone else, and the maintainers are fairly well versed > in the Bug system; merging bugs is not all that hard. Yeah, I just got finished merging several duplicate bugs for emacsen-common. > Yes. It is a good idea. It just should not be mandatory. OK, I'll buy that. We should just *suggest* it then. -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug reporting proceedure, was Re: Bug#24066: libc6: rsh segfaults as , a result of new libc 2.0.7r2
Dale Scheetz <[EMAIL PROTECTED]> writes: > What I am requesting is that the submitter of a bug take some time, in > exchange for the time they expect from the maintainer, and verify that the > bug has not been reported already. If it has, it is appropriate to send > the maintainer a confirmation that you also experienced the bug and any > additional information you can contribute to the solution. This > confirmation should be CC'd to the original bug report, for continuity > purposes. But creating an additional report is both confusing and > administrativly ugly. This should be mandated (I thought it was). Not that I've always been as careful as I should have been, but I've tried to be pretty good. It's extremely helpful (especially for high-traffic-area/critical packages like X and libc) that you not end up with floods of redundant bugs. > I am not looking to stifle information flow about bugs. I am suggesting > that if the reporter of the bug will spend some time "doing the filing" > correctly the task becomes possible and the end user and the maintainer > both benefit. Not only that, but the user might discover a workaround in the bug log that helps them out. This might be the best way to "sell" the idea :> -- Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A094 532B97F5D64E3930 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: HP DAT drives
[EMAIL PROTECTED] (Ben Pfaff) writes: > Does anyone out there have experience with Hewlett-Packard SCSI DAT > drives, specifically the 12/24 GB models? I'm thinking about getting > a backup system for my computer (which has 12 GB of HDDs), and this > seems to be the best deal out there. Do they cooperate nicely with > other devices (HDDs) on a SCSI bus? Does `dump' work okay with them? > How loud is one of them; would it wake up me or my roommate when it > kicks in for a backup at 3:00 a.m.? DAT drives make very little noise if you keep the data streaming to them as fast as they can handle it. If you don't they may make clicking noises every now and then. I'd highly recommend you look into amanda. I use it here, and as long as you're willing to spring for more tapes than you probably originally intended, it does a wonderful job. I also recommend turning off the drive's hardware compression and letting amanda handle it with gzip (unless CPU time is a big deal). -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: New ae uploaded to Incoming
Raul Miller <[EMAIL PROTECTED]> writes: > C-? is fine for those environments where it's not DEL, but the proper > way to implement help has all unbound keys suggest the cannonical help > key (for example: "press F1 for help"). Ok, but Dale's trying to find control bindings for everything, in addition to the function keys, I believe for terminals where the function keys don't work right. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: New ae uploaded to Incoming
Dale Scheetz <[EMAIL PROTECTED]> writes: > I did the best I could at replacing all the "critical" functionality with > control keys so that ae can function in both a dumb terminal as well as > the full featured console. Because of the limited choices the alphabet > provides, I was not able to convert all the keys to control keys, so the > block, cut, and paste functions, the help key, undo, and case flip are > still controled by the function keys. I am taking suggestions for control > key sequences for these functions, and will do what I can to convert all > of them in the near future. Well, if you're emulating emacs, I'd thing that mapping block -> C-space cut-> C-w paste -> C-y undo -> C-u (not really right, I know...) might make sense. And what about C-? for help? All this predicated on the fact that I know exactly none of the restrictions on the choices.. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Deity looks AWESOME
Bob Nielsen <[EMAIL PROTECTED]> writes: > I thought it was an additional method and dselect was going to remain for > the graphically-impaired (IIRC deity runs under X). It has a text mode as well. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: simple smail questions
Otavio Exel <[EMAIL PROTECTED]> writes: > - I saw a message here stating (not in these words) that "smail is dead; > use exim instead"; is it true? I can see that smail is still > (according to Debian) the "reccomended MTA for Debian"! note that I'm > not spam-asking which one you use! :-) I have no idea if smail is "dead", but there has been some discussion about making exim the default mailer. Personally, I use exim. I started with sendmail, switched to qmail, realized that qmail was likely to stay non-free, and then switched to exim, which has worked quite well. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: using SUN's monitor in PC ?
"Jens B. Jorgensen" <[EMAIL PROTECTED]> writes: > There's also a site which tells you > how to build a circuit to allow you to drive it here: > > http://cvs.anu.edu.au/monitorconversion/sun.html Is there similar information available on the feasibility of using an NCD Xterminal's monitor on a PC or a PC's monitor on an NCD Xterminal? Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: FTP installing (Hamm)
Mike Patterson <[EMAIL PROTECTED]> writes: > The Problem: Once I get into the actual system, it seems to have ignored my > network settings (ifconfig and route both return just loopback > settings) and /proc/devices doesn't have an eth device listed. So > I try to run pppd-- apparently my old scripts won't work as it doesn't > recognize the command "+ua" anymore. The last bit I know. >From /usr/doc/ppp/README.gz * Pppd options beginning with - or + have been renamed, e.g. -ip became noip, +chap became require-chap, etc. The old options are still accepted for compatibility but may be removed in future. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Exim/Fetchmail (Re:Was Unidentified subject!)
Mike Acklin <[EMAIL PROTECTED]> writes: > I get an error message from the POP3 server that the name is not > acceptable. It wants a domain name on it. Right, I'm pretty sure this is a DNS config problem. Are you running bind on this machine, or are you just using /etc/hosts? > One of the guys on IRC said this was one of the FAQ's and had me place > "receiver_unqualified_net = 127.0.0.0/255.0.0.0" in my /etc/exim.conf file. I think this is on the right track. If you aren't running bind, what do your /etc/hosts, /etc/nsswitch.conf, and /etc/resolv.conf files look like? This problem can probably be fixed by either changes to one of these files, or by a change to /etc/exim.conf. > Is there somewhere I can find a FAQ or Howto for exim or > fetchmail. I have tried the EXIM home page, but to no avail. The best source for docs on exim is to install the exim-doc package, and then run "info exim". There's also a great mailing list where you can nearly always get all the help you need. It should be listed in the docs. Also check out /usr/doc/exim/* -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Unidentified subject!
[EMAIL PROTECTED] writes: > Thinking I wanted to see if the sendmail package now has a better install > script, I have now uninstalled smail, and installed sendmail. I think I am > over my heard. If you're going to move from smail to anything else at the moment, I'd recommend exim over sendmail unless you're already really comfortable with sendmail. There's even been some noise recently about Debian switching from smail to exim as the default mailer. With respect to your fetchmail problems, holler again if exim doesn't fix the problem, and I'll see what I can suggest. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Linix
SKILLS123 <[EMAIL PROTECTED]> writes: > Hi, I am new to this Linix thingy I am interested in Debian, > because I heard its very good. Yes. > I want to know, can I have Win95 and > Debian on my computer? Yes. > And how, do I install and download the file(s)? Please advise... > Thanks, Matt ftp ftp.debian.org/pub/debian/dists/frozen/main/disks-i386/install.txt or ftp ftp.debian.org/pub/debian/dists/stable/main/disks-i386/install.txt I think those pointers are correct. Alternately you can check out the web pages at www.debian.org. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: smail help required [urgentish]
Hamish Moffatt <[EMAIL PROTECTED]> writes: > Will exim support different alias files for different domains? I've > got the hideous smail method for doing that installed -- duplicating > the configuration files in another directory and using dodgy > transports. This is an important requirement here. I believe it will. You just need multiple directors, each one handling a different domain. exim will take the answer from whichever director has one. From the info pages: * If an address is local, it is passed to each configured director in turn until one is able to handle it. If none can, the address is failed. Directors can be targeted at particular local domains, so several local domains can be processed entirely independently of each other. Is that what you meant? -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mail and mailer questions
Stephen Carpenter <[EMAIL PROTECTED]> writes: > I am interested in looking into a new mail setup and I am looking > for some information and recomendations My highest recommendation is Gnus. It's a fantastic mail program that runs inside emacs, and handles news, so most of what you learn for mail is applicable for news or for more general emacs editing purposes. It probably has more features than you could (would want to) ever use. Gnus easily handles threading your mail and news, and splitting your incoming mail into different groups (mailboxes) based on whatever criteria you like. You'd use fetchmail to download your mail to your machine and then Gnus to deal with it, and I'd recommend using the Gnus nnmail backend which stores each message as a separate file, and is faster than the other backends. There are of course other good mail programs. I used to use mh (and exmh) before I switched to Gnus. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Kernel compilation
Jan Vroonhof <[EMAIL PROTECTED]> writes: > Is fakeroot building the norm in hamm? It is for me. Except that when you're working on a normal package, I tend to do: $ fakeroot debian/rules binary during development, and then $ dpkg-buildpackage -rfakeroot when I'm finishing up. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Kernel compilation
Jan Vroonhof <[EMAIL PROTECTED]> writes: > iii) Doing it this way forces you to be root to do a kernel compile. > > I know this probably more of a problem with the packaging system > itself. However it would be very nice if this could be changed. There > should be no need to be root to compile a kernel, only when installing > it. That's what the fakeroot package is for. $ fakeroot make-kpkg --revision foo.1.0 kernel_image will do what you want. No root required. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Kernel compilation
<[EMAIL PROTECTED]> writes: > Not to mention that I personally run devel kernels for various > reasons. Huh? I'm running make-kpkg compiled versions of the 2.1.79 and 2.1.87 built from the source I downloaded from ftp.kernel.org the other day. What's your point? And when I discovered the chown problem with the 2.1.8(early) kernels, it took me one command: $ sudo dpkg -i /usr/local/src/kernel-images/ kernel-image-2.1.79_1.0_i386.deb and a reboot to fix it. -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Any command to spin down a *scsi* drive?
I know you can use hdparm to set the spin-down period on an IDE drive, but is there a similar command for SCSI drives? Thanks -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: HP SureStore DAT24x6e
Glenn Amerine <[EMAIL PROTECTED]> writes: > Does anyone have this automated DAT changer on their network? And if > so, any troubles backing up Debian boxes with it? From what I can tell > from HP's literature, backup software for NetWare and NT are included, > and then it lists a bunch of Unix systems that are supported via "mtx > autoloader utility". I'm assuming NFS mounts? Check out amanda, it handles full network backups from clusters of machines to tape changers, and probably uses mtx (which is a version of mt that understands multiple tapes). -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: thread support?
[EMAIL PROTECTED] writes: > I haven't found any of the Xlib stuff thread safe either. I'm doing > some development using Xlib and did some basic testing under threads > and found it would hang X-windows or even crash X-windows when > trying to do threaded X-window calls. I have found though that you > can do threaded development using the current non-thread safe X > libraries by carefully organizing your code. I've relagated ONE > thread to do ALL of my X calls, and the other threads to do my other > number crunching. I've had no problems with this and everything > seems to work find. It is a bit difficult though to design the code > just right. You might want to look at /usr/doc/libc6/README.Xfree3.2.linuxthreads.gz after installing libc6. I don't know if the default Debian X packages have this built in yet or not... -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Debian on Dell Latitude Laptop
"Eloy A. Paris" <[EMAIL PROTECTED]> writes: > David Engel <[EMAIL PROTECTED]> wrote: > > : Has anyone successfully installed Debian 1.3.1 on a Dell Latitude XPi > : laptop? If so, what it the secret? When I try booting from the > : install/rescue disk, it gets about halfway through "Loading linux..." > : and then reboots. > > OK, I tried today a stock 1.3.1 rescue disk on a Latitude XPi CD > and it hung while loading the md driver. I'll try compiling > a custom kernel with just the stuff I have. > > Somebody else that contacted me in private mail had the same problem. > However, none of us had the problem of the computer rebooting while > loading the kernel. Sorry this is so late, but I haven't had time for debian-user much recently. Note that this problem is usually caused (on laptops) by a bzImage. If the kernel is recompiled as a zImage, you should be fine. I sucessfully used the tecra boot disk (look in the standard Debian disks directory) on one machine (not a tecra) with this problem. Ignore this if it's no longer relevant... -- Rob Browning <[EMAIL PROTECTED]> PGP fingerprint = E8 0E 0D 04 F5 21 A0 94 53 2B 97 F5 D6 4E 39 30 -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: serial port speed
Lawrence <[EMAIL PROTECTED]> writes:
> Anyone knows the default serail port speed? It is 38,400bps? Which
> file responsible for this setting? I want to increase it to 115,200bps.
Most versions of unix only support specifying the modem speed up to
38400 (for historical reasons). So to specify a higher speed, you
have to tell the OS to use a higher speed when something like pppd
asks for 38400. This is accomplished with setserial, see the man page
for details, but you're looking for the spd_hi and spd_vhi options.
I put these commands in an /etc/rc.boot/0setserial-local file so that
they happen every time I boot.
# !/bin/sh
# File: /etc/rc.boot/0setserial-local
STD_FLAGS="spd_vhi"
SETSERIAL=/bin/setserial
echo -n "Local serial port config"
${SETSERIAL} -b /dev/ttyS0 ${STD_FLAGS}
${SETSERIAL} -b /dev/ttyS1 ${STD_FLAGS}
echo "done."
${SETSERIAL} -bg /dev/ttyS{0,1}
--
Rob
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: Transfering system directories to new HD
"C.L. Daugaard" <[EMAIL PROTECTED]> writes: > I've seen (and kept) posts on how to transfer a system to a new HD. > what I'm stumped on is how this is done when /, /usr, and /home are on > *separate partitions* and I want to keep it that way. Can anyone tell > me how this is done? At this state the "find . -mount -depth > -print|cpio -pdmv /newtempmount" method sounds like the most promising, > but how this is done per partition is a mystery. > > My thanks to anyone who can help. As root make a dummy directory in /, and mount all your new partitions under /dummy, so you have /dev/hdb1 23423 1104811166 50% /dummy /dev/hdb2 956015 749859 156765 83% /dummy/usr /dev/hdb3 956015 749859 156765 83% /dummy/home then do (as root) cd / cp -a `ls | grep -v proc` dummy You'll still have to make the proc directory on your new root partition by hand, but the rest should be exactly the way you want. Then you just need to modify /dummy/etc/lilo.conf and /dummy/etc/fstab to reflect the new partition layout, use cfdisk to set the bootable flag as appropriate, and run "lilo -r /dummy". I think that's it, but this is off the top of my head, so be careful. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Transfering system directories to new HD
Rob Browning <[EMAIL PROTECTED]> writes: > cp -a `ls | grep -v proc` dummy Oops, this needs to be cp -a `ls | fgrep -v proc | fgrep -v 'lost+found'` dummy or something similar. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Please recommend a quality 4GB hard drive
Terrence Brannon <[EMAIL PROTECTED]> writes: > We would like a SCSI 4GB hard disk to connect to our PC running Debian > Linux 1.3. We may occassionally want to connect it to our Alpha's running > Redhat Alpha 4.0. > > We want reliability first. Then cost second. I've had really good luck with the latest Seaagte Barracudas. They're very fast, but quite expensive. I've also heard that their cheaper drives might be troublesome. IBM also apparently has some really good drives now. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: XFree86 and Threads...
Dale Martin <[EMAIL PROTECTED]> writes: > I was just curious - since libc6 is thread safe, and GUIs seem to be > something that can be "naturally" multithreaded, is XFree86 > multithreaded under Linux? (or any other system, for that matter?) > > I'm about to get a second PPro for my box at home, and I already have > SMP at work, so the prospect of a multithreaded X is quite interesting > to me. Now that xlib6g and xlib6g-dev have been released (thanks Mark), the answer should be yes, but I haven't tried it yet. Note that you need to be running unstable, and you have to complile all your code with -D_REENTRANT. There's information about X and threads in the book "Programmers Supplement for Release 6", another one of those "X" books. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash [fixed]
Will Lowe <[EMAIL PROTECTED]> writes:
> Yup, kill the "eval" and life is good. Thanks, Mr. Browning.
You're welcome, but sheesh, call me Rob :>
One final word for those who were following this thread.
This
if [ ${PS1:-UNSET} = UNSET ]
should have been
if [ "${PS1:-UNSET}" = UNSET ]
Without the quotes, you can get problems when you have a prompt with
escape characters.
--
Rob
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
"Gonzalo A. Diethelm" <[EMAIL PROTECTED]> writes: > Rob, thanks a lot for your script. I think there is a small glitch, > though: Not surprised :> > > export PS1='\n\!\$ ' > > export PROMPT_COMMAND='eval set_titlebar [EMAIL > PROTECTED]:`my_dirname`' > > When I did this, the title bar showed the full path, without replacing > $HOME with a '~' character. I got rid of the "eval", and things were > fine. Hmm. I have no idea now why that eval's there. I must have thought I needed it at some point (or I copied a pasted something I shouldn't have), but things seem to work fine without it. The eval might also explain why others were getting the "command not found errors", but if so, I don't understand why I wasn't getting them here too. Thanks for pointing that out. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
Will Lowe <[EMAIL PROTECTED]> writes: > Ok, now I get a > [EMAIL PROTECTED]:~: command not found > > [EMAIL PROTECTED]:~ is what should end up in the titlebar, but it's instead > somehow being evaluated. Any other clues? Email me a snippet that causes the problem, and I'll check it out. If a small snippet won't cause it, send me your .bashrc, and the output of 'ls -l .bash*'. Thanks -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
Michael Harnois <[EMAIL PROTECTED]> writes: > Since you're providing enlightenment ... Don't know if I'd go that far :> > This string works find in a "straight" .bashrc. However, when I use it > in your file, it works fine on a login shell. When I start a subshell, > though, I get I'd have to see the file to be sure, but my guess is that it's not the prompt string, but rather a problem with test (i.e. the "[" command). Email me your .bashrc, and send me the result of "ls -l .bash*". -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: IP Masquerading
Mike Patterson <[EMAIL PROTECTED]> writes: > Now I'm working on IP Masquerading (finally!) and in the HOWTO, there is a > command called "ipfwadm". I can't figure out what package it is under (and > it curently doesn't exist on my system) Oh, and you may want netstd, too. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: IP Masquerading
Mike Patterson <[EMAIL PROTECTED]> writes: > Now I'm working on IP Masquerading (finally!) and in the HOWTO, there is a > command called "ipfwadm". I can't figure out what package it is under (and > it curently doesn't exist on my system) It's in the netbase package. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
Will Lowe <[EMAIL PROTECTED]> writes:
> If I do this:
>
> On 5 Sep 1997, Rob Browning wrote:
>
> > set_titlebar () { echo -n "]2;$*"; }
> > export -f set_titlebar
>
> And then someplace call "set_titlebar", I just get ^[]2;$*^G
> echoed to my terminal. It doesn't ever set the titlebar. I'm using rxvt.
> Suggestions?
Uhm, oops, my fault. There were 8-bit characters in that script that
I didn't mention. I don't know what you saw, but what you should have
seen was:
set_titlebar () { echo -n "^[]2;$*^G"; }
Here ^[ and ^G actually have to be real control characters not a ^
followed by a G or [. To type these into a file in emacs, you can use
C-q C-g for ^G and C-q C-[ for ^[. If this doesn't make sense to you,
ask and I'll explain in more detail. You can get the same effect from
a bash prompt with C-q Cv:
quoted-insert (C-q, C-v)
Add the next character that you type to the line
verbatim. This is how to insert characters like
C-q, for example.
Let me know if this doesn't help.
--
Rob
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
Lawrence <[EMAIL PROTECTED]> writes: > Is it possible to do it in tcsh? Good question. I've never used tcsh, but I would guess so. The key thing to note is this escape sequence which changes the titlebar: echo -n "]2;$*" or with a fixed string for illustration: echo -n "]2;My new titlebar goes here" The rest of the stuff is just figuring out what string to put in the titlebar (via $*), and how to hook the prompt change into the right place so that it happens at every prompt. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
"Jens B. Jorgensen" <[EMAIL PROTECTED]> writes: > Well, you should have RTFMpage, but here's the excerpt you want: That seems a little over-harsh. >a command. Bash allows these prompt strings to be cus >tomized by inserting a number of backslash-escaped special >characters that are decoded as follows: [...] This list of bash escapes doesn't tell them the sequence of characters that has to be sent to an *xterm* to set the title-bar, which is what they really need to know. I had to dig around for a while before I figured that out. Maybe the rxvt man page has been improved... They also have to figure out the BASH_PROMPT variable if they want the "~" cleanup, and several other little tidbits that take quite a while to pull together. I certainly don't begurdge saving someone else the time to figure all that out. Others have certainly done the same for me innumerable times. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Prompt in Bash
Ask and ye shall receive :> This is a cut-down version of my bashrc.
It also shows a trick to get around the problem with some shells
actually being login shells, but not calling .bash_login (i.e. X login
shells). I just symlink my .bash_login to my .bashrc, and let .bashrc
handle figuring out what kind of shell it is (see BASH_INITIALIZED
below).
if [ ${PS1:-UNSET} = UNSET ]
then
INTERACTIVE_SHELL=F
else
INTERACTIVE_SHELL=T
if [ "$TERM" = xterm -o "$TERM" = rxvt ]
then
set_titlebar () { echo -n "]2;$*"; }
export -f set_titlebar
my_dirname () {
if [ "${PWD#$HOME}" != "$PWD" ]
then
echo '~'${PWD#$HOME}
else
echo $PWD
fi
}
export -f my_dirname
export PS1='\n\!\$ '
export PROMPT_COMMAND='eval set_titlebar [EMAIL
PROTECTED]:`my_dirname`'
else
export PS1='[EMAIL PROTECTED]:\w\n\!\$ '
fi
fi
if [ "${BASH_INITIALIZED}" != "Yes" ]
then
export BASH_INITIALIZED=Yes
# Do other stuff here that should only happen once, and be
# inherited by sub-shells.
fi
--
Rob
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: Undefined reference to '...'
Jason Gunthorpe <[EMAIL PROTECTED]> writes: > I hope egcs fixes alot of problems like this :< I believe it does, and again, you can solve all these problems with the existing gcc and the repo patch. One thing I didn't mention before was that you do need to compile your code with -frepo. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: SCSI Host Adapter
Steve Witt <[EMAIL PROTECTED]> writes: > I'm hoping that someone with some SCSI experience can give me some > advice here. My general impression is that the three best manufacturers to consider are Buslogic, Adaptec, and for really serious SCSI, DPT. I've always used Adaptec, but that's just because that's what came with the machines I've used. I've heard that the Buslogic cards are as good or better, and they probably deserve Linux patronage more than Adaptec as they've supposedly been much more open to requests for information from Linux driver developers. I've also heard that the Adaptec's are much more sensitive to cable lengths and qualities. I have had some cabling issues, but I couldn't say if it was the adapter's fault. Buslogic's not carried by too many stores, but it shouldn't be hard to find online. Check www.cdw.com, www.megahaus.com, dirt cheap drives, or the drive outlet center (just off the top of my head). I'm not particularly advocating any of these, but they're places to start. > I finally bought a PC because of Linux, the Mac Unix options weren't > coming along fast enough for me. Same here. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Undefined reference to '...'
I found the pointer to the template repository patch for gcc 2.7.*. It's ftp.cygnus.com/pub/g++/gcc-2.7.*-repo.gz. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: .Xdefaults vs. .Xresources
Christopher Ray Martin <[EMAIL PROTECTED]> writes: > I am a little confused as to the difference between these two files. It > seems that they are basically the same, or at least they contain the same > type of information. > > Which is preferrable to use? Can I use one, and then "source" it from the > other file (for programs which may use the other one)? Well, it may not be strictly correct, but my .Xdefaults is a symlink to .Xresources. Works for me. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Undefined reference to '...'
Gilbert Laycock <[EMAIL PROTECTED]> writes: > I believe that this (and many other) problems have been fixed for > gcc/g++ 2.8 which is apparently nearly-but-not-quite-ready for release > (and has been for some time now). You can also solve this problem in 2.7.2* by using the repo patch (available at cygnus' ftp site (don't have a pointer handy)). We've been using it here for quite some time. The only problem I've noticed is that in some cases, it can make g++ consume an inordinate amount of memory. This patch has the added advantage of making sure that the object code for a given template instance is only instantiated once, rather than in each object file where it's mentioned. I've heard that 2.8 solves this problem in a much better manner. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Monitor power saving (seems to be a FAQ).
Will Lowe <[EMAIL PROTECTED]> writes: > On 1 Sep 1997, Rob Browning wrote: > > Where do you do this? Bash_profile? > > #!/bin/sh > > # File: /etc/rc.boot/0vc-powersave > > # Turn on power-saving on the VC's Note: # File: /etc/rc.boot/0vc-powersave So it's in /etc/rc.boot/0vc-powersave, hence it happens every time the machine boots. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: [OFF TOPIC] sig 14?
Nathan E Norman <[EMAIL PROTECTED]> writes: > where can I find a list of which signal numbers correspond to which > error? /usr/include/asm/signal.h -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Monitor power saving (seems to be a FAQ).
There are really 2 issues: getting X to do power saving, and getting the text console to do power saving. This is how I set all my consoles to all have powersaving enabled: #!/bin/sh # File: /etc/rc.boot/0vc-powersave # Turn on power-saving on the VC's test -f /usr/bin/setterm || exit 0 echo -n "Turning power-saving on for VC's:" for i in 1 2 3 4 5 6 do setterm -powersave < /dev/tty$i setterm -blank 15 < /dev/tty$i echo -n " $i" done echo "." exit 0 And this is how I get X to have powersaving enabled: from /etc/X11/XF86Config Section "Device" ... Option "power_saver" ... EndSection >From the end of my /etc/X11/xdm/Xsetup_0 xset dpms 600 900 The command used to be "xset s power...", so if you have an older version of xset than I do then you might need the following instead: # Turn on the builtin screensaver xset s 480 # and set the monitor shutdown time xset s power 900 900 That should do it. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: "mirror" regets whole Debian archive
You could have avoided the transfer with a "mirror -T". Whenever I see it going haywire (usually because I did something stupid to my clock), I stop it and try that first. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Which PGP should I get?
"Gonzalo A. Diethelm" <[EMAIL PROTECTED]> writes: > Should I get pgp-us or pgp-i? I'm living in Chile, if that makes any > difference. > > >From the looks of it, I'd say pgp-i, but just want to make sure... pgp-i is better for those outside the US. It'd be better for those inside the US too if it wasn't for the RSA patent infringement issues... -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: XDM
Joey Hess <[EMAIL PROTECTED]> writes: > Edit /etc/X11/xdm/Xservers: > > :0 local /usr/X11R6/bin/X -bpp 16 Alternately you can put a DefaultColorDepth 16 entry in the appropriate Screen section in your /etc/X11/XF86Config. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Linux in Wired
[EMAIL PROTECTED] (Ted Harding) writes: > However, I'd be most interested to hear of experience with Applixware > (and what about StarOffice?). I used StarOffice here to convert a Word6 document to html so that I could print it. It did a pretty good job, but it was a small document. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Linux in Wired
Jan Vroonhof <[EMAIL PROTECTED]> writes: > Another example is the ease with which you can make flow text around > a picture with a complicated contour in say PageMaker compared to > how difficult that is in TeX. Of course true lumberjacks would just write raw postscript :> I've succesfully used a Mac version of FreeHand (don't recall which version) running under Executor/Linux (www.ardi.com) for this sort of thing. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Linux in Wired
"Marc W. Brooks" <[EMAIL PROTECTED]> writes: > I would think that XFig, gnuplot and LaTeX could produce all the > "flashiness" that one would need. Also, for those not familiar with it, jgraph can produce some really nice, clean postscript graphs from a fairly straightforward input specification. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: upgrade path
"Daniel J. Mashao" <[EMAIL PROTECTED]> writes: > problems. 1.2 worked fine, but I upgraded to 1.3.1 and now my monitor does > not do 800x600. Did you try su mv /etc/X11/XF86Config /etc/X11/XF86Config.old # Just want to make sure there aren't any old ones lying around to # confuse XF86Setup. mv /root/XF86Config /root/XF86Config.old XF86Setup I had the new version of X break several X installations here, but using this approach fixed all of them but one. Timing parameters (and other things) had changed, and this fixed them by getting the setup program to start from scratch. The one that this didn't fix I was able to straighten out by hand (it was a laptop that was a pain the first time I set it up too). I even found on several of the machines, I could use options that weren't available before. For example, the laptop was able to do 16bit with a linear frame buffer. Hope this helps. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: [DEBIAN] AARGH PCMCIA network adaptor still not recognised
[EMAIL PROTECTED] (Nico De Ranter) writes: > file so the user can choose between Win95/DOS/Linux (if there is a > version of lilo that let's the user choose from a menu (no not by > pressing tab or whatever first) please let me know, in the mean time > I just don't regard lilo as "user-friendly"). Even if I boot from a > floppy nothing is recognised :-( I believe the latest (not yet a Debian package) version of lilo (2.0) will do this. You can also look at chos, but it's non-free... -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: A very simple question
"Civ Kevin F. Havener" <[EMAIL PROTECTED]> writes:
> I'll take a crack at this one since I've asked and received an answer to
> the very same question.
>
> I use a two-line prompt that tells me who I'm logged on as and at what
> machine on the first line and what is the full path to the current directory.
> Put this in your ~/.bash_profile or in /etc/profile (for system-wide
> default): PS1='<[EMAIL PROTECTED]>\n<`pwd`>\$ '
Here's another option (from my .bashrc) that works with rxvt's (and
xterms?). It dynamically sets the title bar to [EMAIL PROTECTED]:pwd so
you don't have to have it all on the prompt. If you're not using an
xterm/rxvt, it sets the prompt instead. It also does /home/username
-> ~ translation when appropriate.
if [ ${PS1:-UNSET} = UNSET ]
then
INTERACTIVE_SHELL=F
else
INTERACTIVE_SHELL=T
if [ "$TERM" = x"term" ]
then
set_titlebar () { echo -n "]2;$*"; }
export -f set_titlebar
my_dirname () {
if [ "${PWD#$HOME}" != "$PWD" ]
then
echo '~'${PWD#$HOME}
else
echo $PWD
fi
}
export -f my_dirname
export PS1='\n\!\$ '
export PROMPT_COMMAND='eval set_titlebar [EMAIL
PROTECTED]:`my_dirname`'
else
export PS1='[EMAIL PROTECTED]:\w\n\!\$ '
fi
fi
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: Laptop for Linux Debian
David M <[EMAIL PROTECTED]> writes: > If one were to buy a laptop these days what would be a good brand that is > widely supported by Linux (Debian)? We have had good luck here with Dells and Compaqs. They both took some tinkering (especially with the X setups), and we had to buy AcceleratedX for one of the Dells. Best place for you to poke around would be the Linux on Laptops home page. Unfortunately I don't have the pointer handy, but Altavista should be able to find it for you pretty quick. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with exim and fetchmail.
Victor Torrico <[EMAIL PROTECTED]> writes: > In the interim, using fetchmail and exim, I still can only get my > incoming mail to go into the /var/spool/exim/input directory. How do I > get the mail in this directory out to use it in exmh? What steps must I > follow? I'm brain frazzled as of the moment. Hmm, I have no idea why it's putting your mail there. AFAIK it shouldn't be. All your mail should be going into /var/spool/mail/username. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Sendmail dsc file?
Nathan E Norman <[EMAIL PROTECTED]> writes: > Hmm ... I tried a 'dpkg-source -x sendmail*dsc' and got this message: > "dpkg-source: error: tarfile `./sendmail_8.8.7.orig.tar.gz' contains > object (sendmail-8.8.7/FAQ) not in expected directory > (sendmail-8.8.7.orig)" > > What's that about? I think you probably need a newer version of dpkg(-dev). -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Sendmail dsc file?
Nathan E Norman <[EMAIL PROTECTED]> writes: > I want to compile the ismx patch into sendmail ... there is no > sendmail.dsc file at ftp.debian.org. I know I can just untar the > source, but I was going to try to do it "the Debian Way" ... what am I > missing? ftp.debian.org:/debian/dists/unstable/main/source/mail> ls sendm* sendmail_8.8.7-1.diff.gz sendmail_8.8.7-1.dsc sendmail_8.8.7.orig.tar.gz Is that what you're looking for? -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: restarting daemons
George Bonser <[EMAIL PROTECTED]> writes: > killall -HUP inetd While this will work, killall is a little evil, because it can sometimes accidentially kill some other things you weren't expecting. A more precise way to do this would be: kill -HUP `cat /var/run/inetd.pid` Either one will get the jon done, though... -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Swap additions
David M <[EMAIL PROTECTED]> writes: > Could some of you guys show me a step by step procedure on how to setup > an additional 128MB swap file? Or alternatively where I can find this info. Check out "man mkswap". -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with exim and fetchmail.
Thanks for the reply. It turns out that the problem lies with fetchmail. It's been fixed in an upcoming release. You won't need any hacks in your exim.conf file anymore, nor will you need fetchmail's -mda option. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Proper solution to exim and fetchmail problem.
After some further cogitation, I spoke to the fetchmail author, and it turns out that it was a problem with fetchmail. He's fixed it in the next version, soon to be released. Once that's done fetchmail users on sytems running exim shouldn't need the -mda option to fetchmail or any special workarounds in their exim config. Thanks for the assistance. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with exim and fetchmail (solved).
Actually it turns out that using receiver_unqualified_hosts=myhost.mydomain fixes the problem nicely. Assuming no one else sees a problem with this solution, I'm going to suggest it to the fetchmail FAQ maintainers to replace the current -mda "exim -bm %" solution. Thanks -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with exim and fetchmail.
Victor Torrico <[EMAIL PROTECTED]> writes: > mda "exim -bm %s" Thanks for the help, but I had discovered (see one of my previous messages) that I could get it to work with this option. However, I wanted to know why it was necessary. So far everyone who has a working setup is using it, but according to the fetchmail FAQ, this approach is a bad idea: >From the fetchmail FAQ: using an MDA for delivery is discouraged. If you throw those options away, fetchmail will now forward your mail into your system's normal Internet-mail delivery path. Actually, using an MDA is now almost always the wrong thing; the MDA facility has been retained only for people who can't or won't run a sendmail-like SMTP listener on port 25. The default, SMTP forwarding to port 25, is better for at least two major reasons. One: it feeds retrieved POP and IMAP mail into your system's normal delivery path along with local mail and normal Internet mail, so all your normal filtering/aliasing/forwarding setup for local mail works. Two: because the port 25 listener returns a positive acknowledge, fetchmail can be sure you're not going to lose mail to a disk-full or some other resource-exhaustion problem. Using the SMTP delivery (no mda option) works fine with sendmail, so just wanted to make it work with exim. Thanks -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Incorrect /usr/include/linux and /usr/src/linux sym-links
[EMAIL PROTECTED] (Harmon Sequoya Nine) writes: > Hello. A few problems I've encountered on installing the kernel > source and headers. > 1) If the source is installed first, then the headers, the > /usr/src/linux symlink pointer to the headers directory instead of > the source directory. > 2) Also, if the kernel headers or source is installed, then the > /usr/include/linux directory should be a symlink to > /usr/src/linux/include/linux. Neither of these is a bug. Read /usr/doc/libc5/FAQ.gz. Linking the asm and linux include directories into the src tree is no longer acceptable practice. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Problem with exim and fetchmail.
Adrian Bridgett <[EMAIL PROTECTED]> writes: > I believe that exim wants a fully qualified address. Try putting this in > /etc/exim.conf: > qualify_domain = localhost > qualify_recipient = localhost I tried this, and it didn't help. > mda "exim -bm %s " This fixes the problem, but according to the FAQ this is just masking some other problem with the SMTP listener. Any idea how to fix the listener? Thanks for the help... -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: .deb file corruption?
Dave Neuer <[EMAIL PROTECTED]> writes: > When I download debian package files onto my PowerMac and try to > transfer them via PC-formatted floppy to my Linux box, the files are > showing up under linux all starting with "!" (i.e. > "kernel-package_3.28.deb" becomes "!kernal-package_3.28") and I > can't seem to cp them (the cp command gives me a "can't find action > '!"). I'm not sure why you are getting the !. Usually things like that happen because someone's trying to represent characters that were used in the filename on the Mac to something that DOS is happier with. However, the reason they are causing trouble is because the shell you're using (bash) interprets the ! as a special character (see "man bash" for more details). Bash things you are telling it to find some previous command in the command history and execute it. For example if I typed "!54" to bash, it would execute the 54th command in the history buffer. Seeing ! in a file name really confuses bash, so you have to "escape" the character to tell bash, "no I really do mean that the file name has an exclamation point". To do this, use a backslash. Since !'s in filenames are evil, you should get rid of them, so do this: mv \!kernel-package_3.28.deb kernel-package_3.28.deb This will change the name, removing the !. If you want to fix a whole directory of files at once, do this (after "cd" ing to the directory): for file in * do mv $file `echo $file | perl -pe 's/^!//o'` done (That's off the top of my head, so I'd be careful, but maybe you'll get the idea...) > What's going on? This seems to happen whether I download the files as > binary or text files. Someone's renaming them for you. What programs are you using for the download? Note that with suntar, you probably wouldn't have this problem... > BTW, I can't get the Linux box connected via Ethernet (no drivers) > or modem (can't seem to find modem, and can't transfer any > communications programs like minicom to the Linux box to test for > the modem). Why not transfer minicom via floppy? (or is that a problem because of the "!" problem I was just discussing?) -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: hello - probs
Lindsay Allen <[EMAIL PROTECTED]> writes: > Is hello_1-13.deb still current? I am trying to learn the gentle art of > package making, but this is what happens with dpkg-source. > > elm# dpkg-source -x /cdrom/stable/source/misc/hello_1.3-13.dsc You need to be in the same directory where you are doing the unpacking, so you need to do something like the following: cd ~ cp /cdrom/stable/source/misc/hello_1.3-13.* . dpkg -x hello_1.13-13.dsc > elm# dpkg --build hello-1.3 You probably don't want to be calling dpkg --build directly. Try something like this: cd hello-1.3 su dpkg-buildpackage or cd hello-1.3 su debian/rules build The latter is if you just want to build the .deb package and not the source and diff packages and the changes file. (You could (and probably eventually should) replace the "su" step with the -rsudo (or something similar) option to dpkg-buildpackage, but I didn't want to make things any more complicated than this to start). > In fact hello contains " debian/" rather than "DEBIAN/". debian/ is correct. The DEBIAN directory is just a temporary directory created during the package build steps which are put in motion with dpkg-buildpackage. You should install the dpkg-dev package (if you haven't already) and look at /usr/doc/dpkg/packaging.html/index.html and /usr/doc/dpkg/internals.html/index.html. -- Rob -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Driver installation problems
[EMAIL PROTECTED] writes: > The modem in the PC is an ISA card modem. Will this map to one of > the /dev/ ttyS's? Rember, I'm primarily a Mac person, so I don't > know an IRQ from my left knee, and to me an I/O address is a place > one one of the moons of Jupiter that the postman delivers mail to. Welcome. I was a Mac user for many years. Switched to Debian, and never looked back (although IRQ's are spawn of satan). The hardware can be a pain at times, but the quality of the OS, the support, the performance, the features, and not least of all the philosophy far more than make up for any hardware problems. Note that you generally solve the hardware problems once, and then forget about them. > If Linux can find my modem. Should be able to, unless it's one of those evil WinModems (leaves out some of the chips to save money, and "lets" your CPU handle the job the missing hardware should have been doing). Linux doesn't handle those, but if it's a good old fashioned modem, you should be fine. Prepare for an intro to IRQs and IO ports. I tell you all this because I'm in the mood, and because you *may* need to know it if you have trouble getting the modem to work. IRQ's and I/O ports are system resources, and for the most part you can think of them as a way for hardware to communicate with the rest of the system. For example, a serial port generally needs a hardware address (I/O port) to send data to other devices, and an IRQ (Interrupt ReQuest line) to notify the CPU (or whoever) that it needs attention. A PC has only 15 IRQ's, and these have to be assigned across all the devices that need them. Generally devices cannot share IRQ's, and some devices require more than one. Also some IRQ's are already reserved by the motherboard. It would be nice if the system would just automatically arbitrate who gets which resources without you ever having to muck around with things, but in the PC world, it's not that simple. The PC is an old crufty architecture, so sometimes you have to help it along. Plug and Pray is an attempt to "fix" some of this, but in many cases it's caused as many problems as it's solved. What it boils down to is that for some devices you have to tell them which IRQ they should use, then you have to tell the system about your decision too. Telling the device may mean moving a jumper on the device or flipping a DIP switch, or it may mean going to the machine's BIOS setup screen and selecting some values. Telling the system (at least under linux, if it's necessary (linux can often auto-detect this stuff), usually means running a config program at boot). Most serial ports require one dedicated IRQ, and you tell the serial port what IRQ to use (or put it on automatic) with the BIOS config program. On my computer you get there by pressing F1 during bootup. A PC normally has 2 serial ports, and Debian does a pretty good job of configuring them without any intervention, so you may be able to leave this alone. Unfortunately, the ISA modem card will have it's own on-board serial port, so you have to be a little careful. Basically, you want to make sure (if you can) that it ends up on a different IRQ and port than any other device (including the built in serial ports). You may be able to just set some jumpers on the card to put it where you want it, or you may need to disable one of the other built in serial ports using the BIOS config program so that the ISA modem can use the disabled serial port's now free resources. You can figure out how the serial ports are already configured on your machine (assuming that you have linux up), by running "setserial -bg /dev/ttyS*" as root (see "man setserial" for more info). This will list all the currently configured serial ports and what resources they are using. Here's a sample: # setserial -bg /dev/ttyS* /dev/ttyS0 at 0x03f8 (irq = 4) is a 16550A (spd_vhi) /dev/ttyS1 at 0x02f8 (irq = 3) is a 16550A (spd_vhi) Note one port is using IRQ 4 and one is using IRQ 3. The 0x numbers are the I/O ports they are using. I didn't configure this myself; Debian did it at startup automatically. The script that handles this auto-configuration is /etc/rc.boot/0setserial. If I had your modem installed setserial -bg's output might look like this: /dev/ttyS0 at 0x03f8 (irq = 4) is a 16550A (spd_vhi) /dev/ttyS1 at 0x02f8 (irq = 3) is a 16550A (spd_vhi) /dev/ttyS2 at 0x0??? (irq = 4) is a 16550A (spd_vhi) Note that the ??? would be filled in with some address, but I don't know what it would be. Note also that I depicted two serial ports "sharing" the same IRQ. Things could be set up that way, but you'd have to be sure to never, ever use both ports. The sytem would get confused about "who was saying what". What you'd really like is something like this: /dev/ttyS0 at 0x03f8 (irq = 4) is a 16550A (spd_vhi) /dev/ttyS1 at 0x02f8 (irq = 3) is a 16550A (spd_vhi) /dev/ttyS2 at 0x0??? (irq = 5) is a 16550A (spd_vhi) where ??? is different from the other
